Whether it's Amazon AWS, or Yahoo BOSS, or JournalTOCs, or almost 
anything else -- there are a variety of API's that library software 
wants to use, which require registering an account to use.

They may or may not be free, sometimes they require a credit card 
attached too.

Most of them assume that an individual person is creating an account, 
the account will be in that individual's name, with an email address, etc.

This isn't quite right for a business or organization, like the library, 
right?  What if that person leaves the organization? But all this 
existing software is using API keys attached to 'their' account? Or what 
if the person doesn't leave, but responsibilities for monitoring emails 
from the vendor (sent to that account) change?  And even worse if 
there's an institutional credit card attached to that account.

I am interested in hearing solutions or approaches that people have 
ACTUALLY tried to deal with this problem, and how well they have worked.

I am NOT particularly interested in "Well, you could try X or Y"; I can 
think of a bunch of things I _could_ try myself, each with their 
potential strengths and weaknesses. I am interested in hearing about 
what people actually HAVE tried or done, and how well it has worked.

Has anyone found a way to deal with this issue, other than having each 
API registered to an account belonging to whatever individual staff 
happened to be dealing with it that day?

Thanks for any advice.